Thanks @SteveBowen for the updates here, much appreciated.
We consume several of the streaming endpoints (company profile, filings, charges, officers, insolvency, exemptions and PSCs) for a production pipeline and have been impacted since the outage began. Two questions we’d really value guidance on:
ETA: Any rough estimate for when stream processing will be restored? Even a ballpark window would help us prepare for the backlog surge when it resumes.
Timepoint / resume guidance - Building on @emmanuel.hadoux’s point above about the buffer window: during the outage, connecting with timepoint=00000000 (or with no timepoint at all) returns an open 200 connection but no data. Our concern is that our last
successfully-stored timepoints are now several days old, and as the buffer only holds ~5-7 days of events, we may soon cross the point where a reconnect silently returns nothing instead of a 416.
When the streams come back online:
What timepoint should we use to resume without missing the events that were queued during the outage?
Will the backlogged filings be replayed through the stream with new timepoints once processing catches up, or will we need a snapshot / alternative method to backfill the gap?
A status endpoint (or a distinct response code while streams are down vs. up) would be hugely helpful for handling this programmatically in future - seconding the earlier requests on that.
Hi everyone thanks for your patience. I can confirm we are now back processing so streams and api requests should contain updates. It will take a a day or so for us to come back up to real time.
But it doesn’t seem to be giving us the missed entries from the last week, the data that has come through after the problem has been fixed mainly has DateOfCreation of today 2026-06-17, nvery little from the last 6 or 7 days.
OK - having manually restarted the streams I now see data on all streams. This beggars the question (and I totally understand this outage was a one off in the years I have been consuming these streams) but as all streams have been running (even during the outage) and have had their daily disconnect + re-connect but only the companies stream picked up, is there any technical explanation reason as to why the others never picked up? Do we need to ‘baby-sit’ and manually restart these streams when such an outage occurs?